Despite the rapid advances in elucidating the molecular basis of cancer, an ostensibly more difficult post-genomic challenge will be the functional annotation of signaling pathways and integration of this information into an operational cell-based model. Unfortunately, this is at present challenging, primarily due to the absence of reliable integrative experimental and bioinformatic toolsets to rapidly delineate and describe signaling networks en masse. RNA interference (RNAi) and expression profiling have proven to be extremely potent and versatile experimental tools to identify and validate molecular components of signal transduction pathways. Despite these successes, high-throughput (HT) RNAi screening and expression profiling are technically challenging and significant limitations in the data analysis, integration, and modeling exist. To address these issues, and to expand on previous program funding, we have developed a novel combined experimental and bioinformatics platform to construct, validate, and model cancer-specific signaling pathways on a genome-wide scale. The ultimate goal of the proposed project is to develop and make commercially available an integrative knowledge database with predictive models of signaling networks specific for proliferation and survival of prostate cancer cells, and software tools for analysis and use of this information in the drug discovery process. Under Phase I funding in collaboration with Roswell Park Cancer Institute, we initially propose to develop and commercialize the HT technology for construction of signaling pathways based on loss-of-function RNAi screening with second generation functionally validated (FV) lentiviral shRNA libraries and validation of key regulators and signaling mechanisms by expression profiling analysis in a panel of pathway-specific shRNA- mediated knock-down cell lines. We propose to use our novel HT RNAi resource to delineate the processes which underlie deregulated proliferation in prostate cancer cells. Then, in conjunction with our bioinformatics collaborators at Ariadne Genomics, Inc., we will compare and combine our findings with the data collected from scientific publications and develop a publicly available knowledge base prototype, and in silica models of signaling networks specifically involved in the control of proliferation and survival of prostate cancer cells. Integrative research combining predictive computational models with heterogeneous experimental data have the potential to greatly simplify validation of tissue-specific signaling networks and significantly impact the molecular dissection of human tumorigenesis mechanisms. This research harbors considerable promise to identify new targets for therapeutic intervention, as well as the development of increasingly relevant paradigms for drug discovery. As a result, we foresee that these toolsets will significantly improve the efficiency, economy, and ease of elucidating and modeling of signal transduction networks that drive neoplastic transformation, and will provide basic researchers with preferred, cost-effective alternatives to existing commercially available reagents and software. PUBLIC HEALTH RELEVANCE: The ultimate goal of the proposed project is to develop and make publicly available new, powerful bioinformatics research tools: a knowledge base of functionally validated pathways specific for and controlling prostate cancer cell viability, and supporting software tools for data analysis and prediction of anti-prostate cancer drug targets. The next generation of genetic screening technology based on functionally validated shRNA libraries will be used to generate data necessary for reconstruction of signaling networks in the prostate cancer cell model. The developed bioinformatics tools and technologies will significantly improve the efficiency of Integrative Cancer Biology translational research related to molecular dissection of diverse human cancer mechanisms, improvement of drug discovery research, and therefore, has major implications for the development of new pharmaceuticals. [unreadable] [unreadable] [unreadable]